Lean spot, cannot resolve

[Delox]

Hey everyone,

So my car (1991 Z32) has developed a lean spot I can't seem to tune out. During WOT, there is a spot from around 3800-4500 where AFR jumps up to around 16:1 or so, then when it finally moves over a single column on the fuel map, it blasts forward and stops hesitating, the AFR drops down to 11:1.

I point out that the issue passes as the car enters a new column on the fuel map, because it is consistently resolved right as it moves from the 12th to the 13th "load" column. I tried adjusting the fuel map cells, but regardless bumping up the cells values, it still stays around the same AFR.

I've attached a log, and my current tune for anyone who's got the time to offer some insight!

As far as physically, I've recently replaced the CAS, TPS, Fuel pump, and FPR.

[Delox]

An earlier log and tune, to show the lack of change even with adjusted fuel cells.

[Delox]

bump.

I ought to mention that all those physical changes to the car were done for other reasons, but none of them resolved this issue.

I will say, it does feel healthier with the bump in fuel, so it is slightly responding, but I'd have to really push the values up to cancel out whatever is happening.

I did beat a Gallardo the other night though, that was nice

[Eric]

Try setting the high gear coefficent to zero.

(mandatory if you have a non standard TT diff. )


>because it is consistently resolved right as it moves from the 12th to the 13th "load" column.

what's with the difference in TP-scaling of your fuel/ignition maps ?
Best to keep these in the same range and similar.

Now the 12th/13th row in your fuelmap (between 46-50TP) is the last 2 rows in your ignition map

-Eric

[Delox]

I will try a tune based on your observations and will report back, thank you.

The reason for the difference between the scaling is because I adjusted the scaling to keep the active cells within the total matrix (aka not slamming into the last cell too early, or not utilizing the whole map). I cannot honestly recall if I adjusted the timing scale too much, as it seemed within reason, and also by not scaling it in unison with the fuel map, I kept the timing lower overall (avoiding preignition).

Was that poor logic? This is the first engine I've tuned, so it has been an ongoing learning experience. I imagine if I scale the timing to the same as the fuel, I'll have higher timing (BTDC) values across the board. I already hear a bit of preignition on full throttle pulls by the top end (hence why the tuning is ongoing!) and more timing scares me.


As far as the high gear coef, with this set to zero, it will not access those maps, correct? If I don't adjust this to zero, when does it access those?

Thanks for any insight.

[Delox]

Unfortunate news.

I adjusted (matched) the scaling from the fuel map to the ign map, adjusted the gear coef to 0. I'm glad to know that I'm only accessing one fuel and ign map, I feel more in control now.

However the issue remains (exact same AFR jump) at the same point. I also had to immediately swap the ign scale back, as the preignition was killer (starting at the exact same time the AFR drop occurs). I couldn't even go through with a full pull for datalogging. SO I will try to match the scaling, but will need to pull the timing back hard.

I ran 2 runs, 3.1.4 with all the suggestions you made. then 3.1.5 with the ign scaling reverted back to where it was.

[Delox]

3.1.5 attachments.

[Eric]

Just blindly changing the scales without changing/shifting the corresponding columns is not a very good idea...especially for the ignition map !
You will have to 'shift' the columns (or rows) so they will keep in line with the scales and if necessary fill gaps in the map with averaged values for the time being.

Anyway, it looks like your max. tp at full load is approx. 55 or so.
However your max. tp (last column) in the fuelmap is 74 and the ignition map it is more realistic, at 57
On top of that you haven't changed anything on the O2 feedback in the fuelmap, which is on until 46 TP (in some areas), which may also be the cause of your flat spot (among other problems it will cause leaving it like this).
In the cells that have O2 feedback switched on, the ECU will ignore the values in these particular cells and use the o2 sensor(s) to aim for 14.7 AFR

So my advice is to use the tp-scaling of the ignition map for the fuelmap (you will probably have to redo/shift the fuelmap a little here and there, see my first comment) and disable the o2-feedback everywhere in the fuelmap(s)


>As far as the high gear coef, with this set to zero, it will not access those maps, correct?

correct, the high gear maps are not used when this is set to 0.


-Eric


PS: I don't see the flat spot anymore in your latest logfiles ? (...in the previous files you could see it quite clearly)

[Delox]

Firstly, thanks for your time. I appreciate it vastly!

(I am not at home to have nistune open, so I am going from memory for this response)

So, I need to make sense of the TP scaling. You want to have it scaled so you do not exceed your maps max TP column (so at max load, your TP is aligned with your final column, ensuring you utilize your map well). If so, where are you reading my max TP as 55 from? I am set to low boost right now (15-16psi), but I recall hitting the final column when I had it at higher boost (18-19psi).

As for shifting/filling in gaps. Say my max TP was 65, and I wanted to scale it so my current last column (say 55) was shifted over 2 columns to the left. I would want to "chop out" 2 columns somewhere on the left, and average out to missing columns among a few surrounding columns, to ensure smoothness. Then, with the 2 newly opened (blanked) columns, I would create new values which would be appropriate for the increased TP/load. I wouldn't necessarily chop out actual columns, I'm just trying to vocalize shifting/smooshing the existing map and adjusting the newly raised TP scale with appropriate values. Is this sound logic to what you are suggesting?

For the O2 feedback. I know I've been lazy with fixing the low load aspect of the maps. When you say to turn off the feedback, do you mean turn it off, tune with it off, then re-enable it (at least at some lower/cruising cells)?

Finally (again going from memory), the flat spot in the AFR still remained (in 3.1.5, when I was actually able to perform a pull), and I recall the kick when the hesitation ended.

[Eric]

So, I need to make sense of the TP scaling. You want to have it scaled so you do not exceed your maps max TP column (so at max load, your TP is aligned with your final column

correct, best is to give the the last value of the TP-scale about +10% extra on your maximum TP at full load, so you make maximal use of the map.
so say, you reach 60 tp on full load, set the last value in the TP scale to 65 or 66.
even if you reach 75 TP by accident or because of boost-overshoot or something, it will still use the last column in the map.

If so, where are you reading my max TP as 55 from?

It's the max tp on full load in your logfiles

As for shifting/filling in gaps. Say my max TP was 65, and I wanted to scale it so my current last column (say 55) was shifted over 2 columns to the left. I would want to "chop out" 2 columns somewhere on the left, and average out to missing columns among a few surrounding columns, to ensure smoothness. Then, with the 2 newly opened (blanked) columns, I would create new values which would be appropriate for the increased TP/load. I wouldn't necessarily chop out actual columns, I'm just trying to vocalize shifting/smooshing the existing map and adjusting the newly raised TP scale with appropriate values. Is this sound logic to what you are suggesting?

Yes, this is what I mean.
I have attached a file where I did this and also disabled the O2 flags in the areas where it needs to be disabled.
(during cruise the O2 feedback is still used).

You can try it, but give it a check-over before trashing it with 20+ psi, as it's a quick 5 minute modification/rescaling/smoothing of your old/existing maps.



-Eric

[Delox]

Great, that tune cleared up what I was wondering.

Now, I ran the car with that tune, and it feels good. I datalogged some pulls through the gears, and the hesitation and snap to normal AFR is very visible.

You can see the AFR bounce around everywhere from 2500 to 4500, getting especially high at 4300 or so (lean enough to cause hesitation/sputter), then drop to a nice high-11 low-12 AFR.

I also need to bump up the latency on the injectors, as its still lean around idle (with O2's completely disabled, and just above in the area you disabled them)


EDIT: take a look at the injection timing and the TP trends. They ramp up until they finally plateau, the AFR snap occurs.

[Eric]

Ok, that's weird...I can't find anything else in the maps/tables that can cause this.
maybe it is caused by a ROM incompatibility.

What is nistune showing to be your base-image ? (shown in the main bar, 23710-xxxxx REV: xx)

[Delox]

We've actually had this convo before. At some point I had a 48XXX base on there somehow, but that has since been resolved.

Currently: 23710-41P60 Rev: 14 (U)

[Eric]

In that case I'm out of ideas.

maybe there's a corruption in the code section within the Nistune board and it needs a reflash, using a base-image programmer.

In that case I'd suggest you have the Nistune board flashed with a 47P10 base image at the same time, as that particular image seems to be the most stable 8-bit Z32 ROM and it's the one we have most knowledge of.

[Delox]

Damn. I do not believe that's really an easily achievable option for me. I don't know anyone local who would have a base image programmer.

I wonder if Matt has any input.